Study of the dynamics of market depth: the case of the Tunis Stock Exchange

Abstract

This paper aims to estimate a dynamic measure of market depth, called VNET (directional net volume), for a stock listed on the Tunis Stock Exchange using high-frequency data.  VNET measures the difference between the volume of transactions initiated by buyers and sellers during the time required to move the price to a certain number of ticks. It is a multidimensional concept that considers the three facets of liquidity: quantity, price, and time. VNET provides a measure of depth for a given price duration, which can be studied throughout the trading day to understand the short-term dynamics of liquidity. To do this, we model the price duration using an Autoregressive Conditional Duration (ACD) model, which allows us to consider the specificities of very high-frequency data irregularly spaced over time, to obtain an unbiased and efficient measure. The nature of the ACD model makes it possible to predict future variations in the liquidity of a security. By identifying the right time to buy or sell, VNET is a good tool for any optimal trading strategy. The empirical results indicate that the market depth depends on internal trading conditions.

Keywords: Market microstructure, Asymmetric information, Liquidity, Market depth, ACD model.               

JEL Classification: C41, D82, G12  

Paper type: Empirical research